muck-feedbag 0.6.0 → 0.7.1

data/ChangeLog

@@ -1,3 +1,9 @@
+* 0.6 - Fri Mar  5 20:10:33 EST 2010
+  - Added bin/feedbag.
+  - Removed the args[:narrow] option, not really needed.
+  - Handle case where feed URLs contain GET parameters; add tests
+    by Patrick Reagan <patrick.reagan@viget.com>.
+
 * 0.5.99 - Tue May 12 12:52:22 EDT 2009
   - Added rails/init.rb to load easily on a Rails app.
 

data/Rakefile

@@ -0,0 +1,59 @@
+require 'rubygems'
+require 'rake'
+require 'rake/testtask'
+require 'rake/rdoctask'
+
+begin
+  require 'rcov/rcovtask'
+  Rcov::RcovTask.new do |t|
+    t.libs << 'lib'
+    t.pattern = 'test/*_test.rb'
+    t.verbose = true
+    t.output_dir = 'coverage'
+    t.rcov_opts << '--exclude "gems/*"'
+  end
+rescue LoadError
+  task :rcov do
+    abort "RCov is not available. In order to run rcov, you must: sudo gem install spicycode-rcov"
+  end
+end
+
+desc 'Test muck-feedbag.'
+Rake::TestTask.new(:test) do |t|
+  t.libs << 'test'
+  t.test_files = FileList["test/feedbag_test.rb"]
+  t.verbose = true
+end
+
+task :test => :check_dependencies
+task :default => :test
+
+begin
+  require 'jeweler'
+  Jeweler::Tasks.new do |gem|
+    gem.name = "muck-feedbag"
+    gem.summary = "Fork of the feedbag gem."
+    gem.description = "This gem will return title and url for each feed discovered at a given url"
+    gem.email = "justin@tatemae.com"
+    gem.homepage = "http://github.com/tatemae/muck-feedbag"
+    gem.authors = ["Axiombox", "David Moreno", "Joel Duffin", "Justin Ball", "Fabien Penso"]
+    gem.add_development_dependency "shoulda"
+    # gem is a Gem::Specification... see http://www.rubygems.org/read/chapter/20 for additional settings
+  end
+  Jeweler::GemcutterTasks.new
+rescue LoadError
+  puts "Jeweler (or a dependency) not available. Install it with: sudo gem install jeweler"
+end
+
+Rake::RDocTask.new do |rdoc|
+  if File.exist?('VERSION')
+    version = File.read('VERSION')
+  else
+    version = ""
+  end
+
+  rdoc.rdoc_dir = 'rdoc'
+  rdoc.title = "muck-feedbag #{version}"
+  rdoc.rdoc_files.include('README*')
+  rdoc.rdoc_files.include('lib/**/*.rb')
+end

data/VERSION

@@ -1 +1 @@
-0.6.0
+0.7.1

data/bin/feedbag

@@ -0,0 +1,28 @@
+#!/usr/bin/env ruby
+
+require "rubygems"
+require "feedbag"
+
+def usage
+  %Q{
+    #{$0} <url 1> [<url 2> <url 3> ... <url n>]
+  }
+end
+
+if ARGV.empty?
+  puts usage
+  exit 1
+end
+
+ARGV.each do |url|
+  puts "== #{url}:"
+  feeds = Feedbag.find url
+  if feeds.empty?
+    puts "   no feeds found!"
+  else
+    feeds.each do |f|
+      puts " - #{f}"
+    end
+  end
+end
+

data/feedbag.gemspec

@@ -2,19 +2,22 @@
  
 Gem::Specification.new do |s|
   s.name = %q{feedbag}
-  s.version = "0.5.103"
-  s.homepage = "http://axiombox.com/feedbag"
-  #s.rubyforge_project = "feedbag"
+  s.version = "0.6"
+	s.homepage = "http://axiombox.com/feedbag"
+	s.rubyforge_project = "feedbag"
  
   s.authors = ["Axiombox", "David Moreno"]
-  s.date = %q{2009-02-10}
+  s.date = %q{2010-03-05}
   s.description = %q{Ruby's favorite feed auto-discoverty tool}
   s.email = %q{david@axiombox.com}
   s.extra_rdoc_files = ["README.markdown", "COPYING"]
-  s.files = ["lib/feedbag.rb", "benchmark/rfeedfinder_benchmark.rb"]
+  s.files = ["lib/feedbag.rb", "benchmark/rfeedfinder_benchmark.rb", "bin/feedbag"]
   s.has_rdoc = true
   s.rdoc_options = ["--main", "README.markdown"]
   s.summary = %q{Ruby's favorite feed auto-discovery tool}
 	s.add_dependency("hpricot", '>= 0.6') 
+	s.bindir = 'bin'
+	s.default_executable = %q{feedbag}
+	s.executables = ["feedbag"]
 end
 

data/lib/feedbag.rb

@@ -21,9 +21,10 @@ require "hpricot"
 require "open-uri"
 require "net/http"
 require 'timeout'
+require 'iconv'
 
 module Feedbag
-  Feed = Struct.new(:url, :title)
+  Feed = Struct.new(:url, :title, :human_url)
 
 	@content_types = [
 		'application/x.atom+xml',
@@ -41,6 +42,8 @@ module Feedbag
 		# use LWR::Simple.normalize some time
 		url_uri = URI.parse(url)
 		url = "#{url_uri.scheme or 'http'}://#{url_uri.host}#{url_uri.path}"
+		url << "?#{url_uri.query}" if url_uri.query
+		
 		# hack:
 		url.sub!(/^feed:\/\//, 'http://')
 
@@ -69,31 +72,34 @@ module Feedbag
     return self.add_feed(url, nil) if looks_like_feed? url
 
 		# check if feed_valid is avail
-    unless args[:narrow]
-      begin
-	  		require "feed_validator"
-		  	v = W3C::FeedValidator.new
-			  v.validate_url(url)
-  			return self.add_feed(url, nil) if v.valid?
-	  	rescue LoadError
-		  	# scoo
-  		rescue REXML::ParseException
-	  	  # usually indicates timeout
-		    # TODO: actually find out timeout. use Terminator?
-		    $stderr.puts "Feed looked like feed but might not have passed validation or timed out"
-      rescue => ex
-	  		$stderr.puts "#{ex.class} error ocurred with: `#{url}': #{ex.message}"
-		  end
-    end
+    begin
+  		require "feed_validator"
+	  	v = W3C::FeedValidator.new
+		  v.validate_url(url)
+			return self.add_feed(url, nil) if v.valid?
+  	rescue LoadError
+	  	# scoo
+		rescue REXML::ParseException
+  	  # usually indicates timeout
+	    # TODO: actually find out timeout. use Terminator?
+	    # $stderr.puts "Feed looked like feed but might not have passed validation or timed out"
+    rescue => ex
+  		$stderr.puts "#{ex.class} error ocurred with: `#{url}': #{ex.message}"
+	  end
 
 		begin
       Timeout::timeout(10) do
         open(url) do |f|
-          if @content_types.include?(f.content_type.downcase)
-            return self.add_feed(url, nil)
-          end
-
-          doc = Hpricot(f.read)
+          content_type = f.content_type.downcase
+  				if content_type == "application/octet-stream" # open failed
+  				  content_type = f.meta["content-type"].gsub(/;.*$/, '')
+  				end
+  				if @content_types.include?(content_type)
+  					return self.add_feed(url, nil)
+  				end		
+          				
+					ic = Iconv.new('UTF-8//IGNORE', f.charset)
+          doc = Hpricot(ic.iconv(f.read))
 
           if doc.at("base") and doc.at("base")["href"]
             $base_uri = doc.at("base")["href"]
@@ -109,14 +115,20 @@ module Feedbag
             end
           end
 
-          unless args[:narrow]
-            (doc/"a").each do |a|
-              next unless a["href"]
-              if self.looks_like_feed?(a["href"])
-                self.add_feed(a["href"], url, $base_uri, a["title"] || a.inner_html || a['alt']) # multiple fallbacks, first title, then the tag content, then the alt tag (in case of image)
-              end
+          (doc/"a").each do |a|
+            next unless a["href"]
+            if self.looks_like_feed?(a["href"]) and (a["href"] =~ /\// or a["href"] =~ /#{url_uri.host}/)
+              self.add_feed(a["href"], url, $base_uri, a["title"] || a.inner_html || a['alt'])
+            end
+          end
+
+          (doc/"a").each do |a|
+            next unless a["href"]
+            if self.looks_like_feed?(a["href"])
+              self.add_feed(a["href"], url, $base_uri, a["title"] || a.inner_html || a['alt'])
             end
           end
+          
         end
       end
 		rescue Timeout::Error => err
@@ -161,7 +173,7 @@ module Feedbag
 		end
 
 		# verify url is really valid
-		$feeds.push(Feed.new(url, title)) unless $feeds.any? { |f| f.url == url }# if self._is_http_valid(URI.parse(url), orig_url)
+		$feeds.push(Feed.new(url, title, orig_url)) unless $feeds.any? { |f| f.url == url }# if self._is_http_valid(URI.parse(url), orig_url)
 	end
 
 	# not used. yet.

data/muck-feedbag.gemspec

@@ -0,0 +1,62 @@
+# Generated by jeweler
+# DO NOT EDIT THIS FILE DIRECTLY
+# Instead, edit Jeweler::Tasks in Rakefile, and run 'rake gemspec'
+# -*- encoding: utf-8 -*-
+
+Gem::Specification.new do |s|
+  s.name = %q{muck-feedbag}
+  s.version = "0.7.1"
+
+  s.required_rubygems_version = Gem::Requirement.new(">= 0") if s.respond_to? :required_rubygems_version=
+  s.authors = ["Axiombox", "David Moreno", "Joel Duffin", "Justin Ball", "Fabien Penso"]
+  s.date = %q{2011-01-31}
+  s.default_executable = %q{feedbag}
+  s.description = %q{This gem will return title and url for each feed discovered at a given url}
+  s.email = %q{justin@tatemae.com}
+  s.executables = ["feedbag"]
+  s.extra_rdoc_files = [
+    "ChangeLog",
+    "README.markdown",
+    "TODO"
+  ]
+  s.files = [
+    "COPYING",
+    "ChangeLog",
+    "README.markdown",
+    "Rakefile",
+    "TODO",
+    "VERSION",
+    "benchmark/rfeedfinder_benchmark.rb",
+    "bin/feedbag",
+    "feedbag.gemspec",
+    "index.html",
+    "lib/feedbag.rb",
+    "muck-feedbag.gemspec",
+    "test/atom_autodiscovery_test.rb",
+    "test/feedbag_test.rb",
+    "test/test_helper.rb"
+  ]
+  s.homepage = %q{http://github.com/tatemae/muck-feedbag}
+  s.require_paths = ["lib"]
+  s.rubygems_version = %q{1.3.7}
+  s.summary = %q{Fork of the feedbag gem.}
+  s.test_files = [
+    "test/atom_autodiscovery_test.rb",
+    "test/feedbag_test.rb",
+    "test/test_helper.rb"
+  ]
+
+  if s.respond_to? :specification_version then
+    current_version = Gem::Specification::CURRENT_SPECIFICATION_VERSION
+    s.specification_version = 3
+
+    if Gem::Version.new(Gem::VERSION) >= Gem::Version.new('1.2.0') then
+      s.add_development_dependency(%q<shoulda>, [">= 0"])
+    else
+      s.add_dependency(%q<shoulda>, [">= 0"])
+    end
+  else
+    s.add_dependency(%q<shoulda>, [">= 0"])
+  end
+end
+

data/test/atom_autodiscovery_test.rb

@@ -1,10 +1,4 @@
-#!/usr/bin/ruby
-
-require "#{File.dirname(__FILE__)}/../feedbag"
-require "test/unit"
-require "open-uri"
-require "hpricot"
-require "pp"
+require File.dirname(__FILE__) + '/test_helper'
 
 class AtomAutoDiscoveryTest < Test::Unit::TestCase
 	def test_autodisc
@@ -20,14 +14,14 @@ class AtomAutoDiscoveryTest < Test::Unit::TestCase
 			f = Feedbag.find url
 
 			assert_instance_of Array, f
-			assert f.size == 1, "Feedbag didn't find a feed on #{url} or found more than one"
+			f.size.should == 1, "Feedbag didn't find a feed on #{url} or found more than one"
 
 			puts " found #{f[0]}"
 			feed = Hpricot(open(f[0]))
 	
 			(feed/"link").each do |l|
 				next unless l["rel"] == "alternate"
-				assert_equal l["href"], url
+				url.should == l["href"]
 			end
 
 			# ahora me voy al siguiente

data/test/feedbag_test.rb

@@ -0,0 +1,19 @@
+require File.dirname(__FILE__) + '/test_helper'
+
+class FeedbagTest < ActiveSupport::TestCase
+  
+  test "Feedbag.feed? should know that an RSS url is a feed" do
+    rss_url = 'http://example.com/rss/'
+    Feedbag.stubs(:find).with(rss_url).returns([rss_url])
+    
+    assert Feedbag.feed?(rss_url)
+  end
+  
+  test "Feedbag.feed? should know that an RSS url with parameters is a feed" do
+    rss_url = "http://example.com/data?format=rss"
+    Feedbag.stubs(:find).with(rss_url).returns([rss_url])
+    
+    assert Feedbag.feed?(rss_url)
+  end
+  
+end

data/test/test_helper.rb

@@ -0,0 +1,11 @@
+require 'rubygems'
+
+require 'test/unit'
+
+
+require 'active_support'
+require 'active_support/test_case'
+
+require 'mocha'
+
+require File.dirname(__FILE__) + '/../lib/feedbag'

metadata

@@ -1,77 +1,102 @@
 --- !ruby/object:Gem::Specification 
 name: muck-feedbag
 version: !ruby/object:Gem::Version 
-  version: 0.6.0
+  hash: 1
+  prerelease: false
+  segments: 
+  - 0
+  - 7
+  - 1
+  version: 0.7.1
 platform: ruby
 authors: 
+- Axiombox
+- David Moreno
 - Joel Duffin
 - Justin Ball
+- Fabien Penso
 autorequire: 
 bindir: bin
 cert_chain: []
 
-date: 2009-11-10 00:00:00 -07:00
-default_executable: 
+date: 2011-01-31 00:00:00 -07:00
+default_executable: feedbag
 dependencies: 
 - !ruby/object:Gem::Dependency 
   name: shoulda
-  type: :development
-  version_requirement: 
-  version_requirements: !ruby/object:Gem::Requirement 
+  prerelease: false
+  requirement: &id001 !ruby/object:Gem::Requirement 
+    none: false
     requirements: 
     - - ">="
       - !ruby/object:Gem::Version 
+        hash: 3
+        segments: 
+        - 0
         version: "0"
-    version: 
+  type: :development
+  version_requirements: *id001
 description: This gem will return title and url for each feed discovered at a given url
 email: justin@tatemae.com
-executables: []
-
+executables: 
+- feedbag
 extensions: []
 
 extra_rdoc_files: 
 - ChangeLog
 - README.markdown
+- TODO
 files: 
-- .gitignore
 - COPYING
 - ChangeLog
 - README.markdown
+- Rakefile
 - TODO
 - VERSION
 - benchmark/rfeedfinder_benchmark.rb
+- bin/feedbag
 - feedbag.gemspec
 - index.html
 - lib/feedbag.rb
-- rails/init.rb
+- muck-feedbag.gemspec
 - test/atom_autodiscovery_test.rb
+- test/feedbag_test.rb
+- test/test_helper.rb
 has_rdoc: true
 homepage: http://github.com/tatemae/muck-feedbag
 licenses: []
 
 post_install_message: 
-rdoc_options: 
-- --charset=UTF-8
+rdoc_options: []
+
 require_paths: 
 - lib
 required_ruby_version: !ruby/object:Gem::Requirement 
+  none: false
   requirements: 
   - - ">="
     - !ruby/object:Gem::Version 
+      hash: 3
+      segments: 
+      - 0
       version: "0"
-  version: 
 required_rubygems_version: !ruby/object:Gem::Requirement 
+  none: false
   requirements: 
   - - ">="
     - !ruby/object:Gem::Version 
+      hash: 3
+      segments: 
+      - 0
       version: "0"
-  version: 
 requirements: []
 
 rubyforge_project: 
-rubygems_version: 1.3.5
+rubygems_version: 1.3.7
 signing_key: 
 specification_version: 3
 summary: Fork of the feedbag gem.
 test_files: 
 - test/atom_autodiscovery_test.rb
+- test/feedbag_test.rb
+- test/test_helper.rb

data/.gitignore

@@ -1,30 +0,0 @@
-*.swp
-**/*.pid
-log/*.log
-log/*.pid
-tmp
-.DS_Store
-public/cache/**/*
-public/system/**/*
-doc/**/*
-db/*.sqlite3
-.project
-.loadpath
-nbproject/
-.idea
-testjour.log
-*.so
-*.o
-Makefile
-mkmf.log
-*.bundle
-conftest
-content/
-.idea
-*.sw?
-.DS_Store
-coverage
-rdoc
-pkg
-pkg/*
-log/*

data/rails/init.rb

@@ -1 +0,0 @@
-require File.join File.dirname(__FILE__), "..", "lib", "feedbag"